Much has been written about engagement and how to measure it – but it’s a difficult idea to pin down.
Engagement is helpful because as activities online and site continue to develop and diversify, so does the need to qualify behavior.
Generally, Web Analysts and site stakeholders use clearly identified goals and a few corresponding “hard” conversion metrics to assess performance, but using a set of trended, softer metrics can help build a more rounded picture of visitor engagement.
Engagement metrics as their most basic are often considered in the context of average volume of pages viewed during a visit and the average amount of time taken to view them. This on its own has its problem. For instance, it doesn’t answer the question whether a visitor was engaged on the site in a positive or negative way.
As with so much in Web Analytics, some context is useful in assessing engagement, when looking at a set of engagement metrics over time. What happens to them when conversion to the desired goal increases? Do visitors that convert spend more or less time on your site? Do they view more or fewer pages that the average visitor? Is the site’s bounce rate higher or lower? How often do they return to the site and at what intervals? In most cases a high bounce rate is bad but in some instances, it may act as a filter, sorting the wheat from the chaff, leaving only the more interested or “engaged” visitors.
Using segmentation to isolate and compare referring source of traffic will also help in understanding which users are happily engaged with your site content and which are less. So engagement metrics for the new and repeat visitors may well differ. Alternatively, creating a segment using the combined benchmark data that’s already been established for positive engagement can help in unearthing top referring sources and keywords to zero in on when developing an acquisition strategy.
Having a conversion benchmark provides a target to aim for; establishing an engagement protocol offers insight into what good and poor visitor behavior look like. But to truly understand engagement we must go beyond just web analytics and look at other sources of data.